This study investigates the behavior of interacting surface cracks at the circumferential weld toe of monopile-supported offshore wind turbines. Relying on a numerical model that explicitly considers weld profiles, we explore the impact of crack interaction and loading scenarios on crack propagation. Our findings reveal that, initially, surface cracks grow independently, resembling single crack behavior. However, a pronounced interaction effect accelerates their growth as cracks propagate further, potentially leading to crack coalescence, high stress intensity factors, and reduced fatigue life. Consequently, this work highlights the need for integrating specific weld geometry representation in numerical models, as neglecting this can lead to significantly inaccurate fatigue life estimates in typical practical applications. Moreover, this study points out the challenge in accessing representative crack growth material parameters, vital for accurately evaluating the fatigue life of structural connections in offshore wind turbines.@en

Aging wind energy assets demand the development of methods able to effectively support informed decision-making. These needs have inspired the use of data-driven methodologies, which offer valuable insights to wind turbine owners and/or operators. Many approaches can be found in the literature for extrapolating fatigue damage measurements to estimate the lifetime of wind turbines. In some cases, resampling approaches are proposed to compute the confidence levels associated with the generated projections, yet a standardized framework has not been adopted. Most reported studies identify the relationship between short-term damage and long-term Environmental and Operational Conditions (EOCs) by mainly rendering mean lifetime predictions and their associated confidence levels, whereas additional predicted lifetime statistical information is usually overlooked. In this work, we showcase the importance of properly accounting for the variability in lifetime predictions, describe how to summarize binned damages using statistical estimators and investigate bootstrapping variants for computing the confidence levels in the generated damage estimators.@en

The application of Deep Reinforcement Learning (DRL) for the management of engineering systems has shown very promising results in terms of optimality and scalability. The interpretability of these policies by decision-makers who are so far mostly familiar with traditional approaches is also needed for implementation. In this work, we address this topic by providing a comprehensive overview of POMDP- and DRL-based management policies, along with simulation-based implementation details, for facilitating their interpretation. By mapping a sufficient statistic, namely a belief state, to the current optimal action, POMDP-DRL strategies are able to automatically adapt in time considering long-term sought objectives and the prior history. Through simulated policy realizations, POMDP-DRL-based strategies identified for representative inspection and maintenance planning settings are thoroughly analyzed. The results reveal that if the decision-maker opts for an alternative, even suboptimal, action other than the one suggested by the DRL-based policy, the belief state will be accordingly updated and can still be used as input for the remainder of the planning horizon, without any requirements for model retraining.@en